Discover, rate, and discuss biology preprints.

This paper introduces LDDM, a generative model for drug design that produces synthetically accessible compounds validated experimentally for five protein targets.

QBayMic improves clustering and feature selection in low-signal microbiome data by using quantum-coupled variational Bayes, outperforming classical methods in stability and accuracy.

The paper introduces a generative model framework that maps protein sequences to microscopy images and vice versa, enabling virtual mutagenesis and functional motif analysis.

The paper presents a comparative mitogenomic and molecular phylogenetic analysis of five agriculturally significant Tephritid fruit fly species in Bangladesh using existing GenBank sequences.

The paper investigates the impact of post-selection inference in scRNA-Seq phenotype testing and proposes batch correction to mitigate false discovery rate inflation.

The paper introduces decemedip, a Bayesian hierarchical model for cell type deconvolution of MeDIP-seq data, showing its accuracy and utility in identifying methylation signatures.

The paper tracks shifts in atrial fibrillation research using fixed-window citation analysis, showing increased focus on pulsed field ablation and precision risk stratification.

The paper introduces MESIC, a method that maps gene expression data into an interpretable semantic space using gene annotations, enabling biological interpretation of clusters and outliers.

LLMsFold designs and validates new small molecules targeting pathogenic proteins using LLMs and biophysical simulations, showing strong in silico binding and drug-like profiles.

This study evaluates two time-series foundation models for cognitive workload classification from eye-tracking data, showing better generalization to unseen individuals than task-specific models.

TRACEDD provides a framework for explainable drug design by integrating multiple agents and tools to coordinate drug discovery tasks from target validation to synthesis planning.

The paper develops a large language model to predict SARS-CoV-2-specific TCRs and shows their correlation with antibody levels post-vaccination.

The paper introduces a two-stage deep learning approach for gene expression analysis in esophageal adenocarcinoma, identifying more expressed genes than traditional methods.

With an incomplete deconvolution reference, operator history changes estimates and associations, and the target is not point-identified; fitdrop scores remedies on shrinkage, coverage, certification and false certification, and prices measurements.

The paper uses a sparse machine learning pipeline to identify multi-omic biomarkers predictive of bronchopulmonary dysplasia in cord blood samples.

DeepSpaceDB 2.0 expands its database to include 1,539 public Xenium datasets, offering improved tools for interactive exploration and analysis of spatial transcriptomics data.

The paper shows that incorporating physics-based priors improves structure-based binding affinity models, achieving performance comparable to ligand-only models on a drug discovery dataset.

This paper evaluates DrugCLIP's contrastive learning embeddings for virtual screening, showing strong performance but limitations under realistic conditions.

DeepVir automates viral dark matter discovery by integrating homology-based alignment and HMM mining, identifying novel viral groups and lineages from bat transcriptomic data.

5,776,417 ColabFold-AlphaFold2 structures cover 98.9% of viral UniProt 2025_03 sequences, 16.4-fold more than earlier versions, with 75.3% high-confidence and 72.6% of viral reference proteomes fully covered.